Relevance Vector Machine Optimization in Automatic Text Summarization

Author

Kania Evita Dewi, S.Pd., M.Si Dr. Ednawati Rainarli, S.Si., M.Si.

Abstrak

This study aims at optimizing the Relevance Vector Machine (RVM) algorithm in automatic text summarization. This research begins by studying various studies on automatic text summarization to find out what features are commonly used in the automatic text summarization process. Each feature value will be calculated as a correlation with the target. The composition of features is determined by obtained correlation value, when the correlation value between features and targets is greater, the feature will take precedence. The results in this study are obtained by using 4 or 6 features that obtain the highest accuracy, which is 55.84%. The conclusion of this study is that the correlation coefficient can be used to determine the order of extraction features.